Malignant astrocytomas are among the most common and deadly brain tumors of childhood. Most affected children die within several years of diagnosis, despite surgical and adjuvant treatment; however, 30 to 40 percent respond favorably to conventional therapy and appear to be cured. The basis for these diverse outcomes has been enigmatic, even taking into account clinical and histological factors. In preliminary studies with an institutional cohort of pediatric malignant gliomas, we hypothesized that molecular markers of tumor biology could supplement histological data to refine prognostic assessments. These studies indicated a strong association between survival and selected markers, particularly TP53 mutation status and MIB-1 labeling index, and suggested that pediatric gliomas of tumorigenesis distinct from their adult counterparts. Based upon these findings, we initiated a more extensive study of the multi- institutional cohort of Children's Cancer Group study CCG-945, the largest group of pediatric high-grade gliomas accrued to date. During the initial funding period of this project (beginning October 1998), archival specimens were obtained from 179 tumors, the vast majority of which were successfully evaluated in initial histological studies of MIB-1 based proliferation labeling and microdissection-based genotyping of TP53 exons 5 to 8. These studies demonstrated a striking association between outcome and both p53 mutation/overexpression status and MIB1 proliferation index, independent of clinical or histological factors, and generated a sizeable resource of microdissected tumor tissue for more extensive analyses. The studies proposed here will employ this unique resource to further characterize the molecular features of tumorigenesis in pediatric malignant gliomas. We hypothesize that categorization of these tumors by their profile of genomic alterations will improve the accuracy of prognostic assessments, and may reveal patterns of abnormalities that are distinct from those of adult high-grade gliomas. To test these hypotheses, we propose studies with the following aims: 1) Determine the incidence of gene alterations affecting p53 function in pediatric high-grade gliomas (i.e., TP53 mutation or deletion, MDM2 amplification, and p14ARF deletion or mutation), and whether these alterations correlate with patient age or therapeutic outcome. 2) Examine the frequency with which G1-S transition control genes are disrupted in these tumors, specifically by deletion of RB1 or CDKN2A, or amplification of CDK4. 3) Assess the frequency and prognostic relevance of alterations in other loci associated with outcome in adult high-grade gliomas, such as PTEN mutation, chromosome 1p and 19q deletions, and EGFR amplification. These markers will be evaluated in the context of other prognostic factors to determine their utility for biologically classifying childhood malignant gliomas. Taken together, this work will provide new insights into the molecular categorization of these tumors, and establish a foundation for risk-adapted therapeutic stratification in future studies of pediatric high-grade gliomas.